Immunity inspired Cooperative Agent based
Security
System
Praneet Saurabh and Bhupendra Verma
Department of Computer Science and Engineering, Technocrats Institute of
Technology, India
Abstract: Artificial
Immune System (AIS) has evolved substantially from its inception and is
utilized to solve complex problems in different domains out of which computer
security is one of them. Computer Security has emerged as a key research area
because of the ever-growing attacks and its methodology. Various security
concepts and products were developed to overcome this alarming situation but
these systems by some means fall short to provide the desired protection
against new and ever-increasing threats. AIS enthused from Human Immune System (HIS)
is considered as an excellent source of inspiration to develop computer
security solution since the previous protect the body from various external and
internal threats very effectively. This paper presents Immunity Inspired
Cooperative Agent based Security System (IICASS) that uses Enhanced Negative
Selection Algorithm (E-RNS) which incorporate fine tuning of detectors and
detector power in negative selection algorithm. These features make IICASS
evolve and facilitate better and correct coverage of self or non-self.
Collaboration and communication between different agents make the system
dynamic and adaptive that helps it to discover correct anomalies with degree of
severity. Experimental results demonstrate that IICASS show remarkable
resilience in detecting novel unseen attacks with lower false positive.
Keywords: Anomaly, human immune system, artificial immune
system, agent.
Received June 3, 2014; accepted May 24, 2016
|
|